This invention relates generally to wafer bonding.
In wafer bonding, two semiconductor wafers may be placed in a face-to-face configuration. A layer on one semiconductor wafer may be transferred to the other semiconductor wafer in a process called wafer bonding. A wide variety of layers may be transferred between semiconductor wafers. One application for wafer bonding is in connection with forming silicon on insulator (SOI) devices.
Generally, a pair of opposed flat silicon wafers are contacted to one another so that they physically and chemically bond. A layer is transferred from a donor wafer to a handle wafer.
One problem with existing wafer bonding processes is that a peripheral region of the handle wafer, generally about 3 to 5 millimeters, may remain unbonded. This unbonded peripheral region is a region on the outer periphery of the wafer extending radially inwardly from the edge of the wafer to a distance of about 3 to 5 millimeters.
As a result of this unbonded region, islands of material, debris, particles, and flakes may collect in the unbonded region created by the resulting edge. These particles may ultimately release, resulting in problematic defects. In addition, the wafers may only have a useable surface area up to 3 to 5 millimeters inwardly of the outermost edge. The unbonded area may result in some loss of useable wafer area.
Thus, there is a need for better ways to wafer bond wafers.